By analogy with an effective model of charged-lepton mass matrix that, with
the inputs of m(e)(exp) and m(mu)(exp), predicts (in a perturbative zero o
rder) m(tau) = 1776.80 MeV close to m(tau)(exp) = 1777.03 (+0.30)(-0.26) Me
V, we construct such a model for quark mass matrices reproducing consistent
ly the bulk of experimental information on quark masses and mixings. In par
ticular, the model predicts \V-ub\ = 0.00313, gamma = - arg V-ub = 63.8 deg
rees and \V-td\ = 0.00785, beta = - arg V-td = 20.7 degrees (i.e., sin 2 be
ta = 0.661 to be compared with the BaBar value sin 2 beta (exp) = 0. 59 +/-
0. 14), if the figures \V-us(exp)\ = 0. 2196, \V-cb(exp)\ = 0. 0402 and m(
s)(exp) = 123 MeV, m(c)(exp) = 1.25 GeV, m(b)(exp) = 4.2 GeV are used as in
puts. Also the rest of CKM matrix elements is predicted consistently by the
experimental data. Here, quark masses and CKM matrix elements (ten indepen
dent quantities) are parametrised by eight independent model constants, wha
t gives two independent predictions, e.g. for \V-ub\ and beta. The consider
ed model deals with the fundament al-fermion Dirac mass matrices, so that t
he neutrino Majorana mass matrix is outside the scheme. Some foundations of
the model are collected in Appendix.